pyopmcsp11: A Python framework using OPM Flow for the CSP11 benchmark project

This repository contains scripts to set up a workflow in python for the three cases in the CSP11 project. Here we use the OPM-Flow simulator.

Installation

You will first need to install

• Flow (https://opm-project.org)

To build dune and the corresponding OPM master branches from source (e.g., you are a macOS user), you can run the script ./build_dune_and_opm-flow.bash, which in turn should build flow in the folder ./build/opm-simulators/bin/flow (for macOS users the dependecies such as boost can be installed using brew or macports).

You can install the requirements in a virtual environment with the following commands:

# Clone the repo
git clone https://github.com/daavid00/pyopmcsp11.git
# Get inside the folder
cd pyopmcsp11
# Create virtual environment
python3 -m venv vpyopmcsp11
# Activate virtual environment
source vpyopmcsp11/bin/activate
# Upgrade pip, setuptools, and wheel
pip install --upgrade pip setuptools wheel
# Install the pyopmcsp11 package (in editable mode for contributions/modifications, i.e., pip install -e .)
pip install .
# For contributions/testing/linting, install the dev-requirements
pip install -r dev-requirements.txt

See the CI.yml script for installation of OPM Flow and the pyopmcsp11 package in ubuntu.

Running pyopmcsp11

You can run pyopmcsp11 as a single command line:

pyopmcsp11 -i some_input.txt -o some_output_folder

Run pyopmcsp11 --help to see all possible command line argument options. Inside the some_input.txt file you provide the path to the flow executable and simulation parameters. See the .txt files in the examples folders.

Running in a docker

Docker container allows to run opm inside an encapsulated environment. To use it either pull the public docker image

	docker pull jafranc/opm-u22-spe11csp

or build it from the source omp-spe11csp-ubuntu.Dockerfile located at ./dockers with command:

	cd dockers
	docker build --build-arg IMG=ubuntu --build-arg VERSION=jammy -t <tag-docker> -f omp-spe11csp-ubuntu.Dockerfile .

Then using docker images, the created or pulled image should be visible. Running the data generator and simulation for a case is then done by

	docker run -e SPE_CASE=<a,b or c> -v <host-path-to-result>:/opt/spe11csp/output_csp11<a,b or c> <tag-docker> 

Then you can inspect what has been produced at <host-path-to-result>. <tag-docker> being the tag you choose when building the image or if pulled from docker:// , jafranc/opm-u22-spe11csp:latest. Eventually <a,b or c> will set a variable in the docker container generating and running the simulation either for csp SPE 11th case a (2D surface conditions), case b (2D reservoir conditions) or c (3D extruded reservoir conditions).

Results can be visualized using ResInsight or other viewers.

Getting started

See the documentation.

About pyopmcsp11

The pyopmcsp11 package is being funded by the HPC Simulation Software for the Gigatonne Storage Challenge project [project number 622059] and Center for Sustainable Subsurface Resources (CSSR) [project no. 331841]. This is work in progress. Here is the link to the csp11 details. Contributions are more than welcome using the fork and pull request approach.